INDEX

Explanations

phrases related to specific events, locations, and titles

oai_token-act-pair · gpt-3.5-turbo

key terms related to societal issues and cultural discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 explorers

-0.56

rients

-0.53

HC

-0.51

orius

-0.49

romeda

-0.48

roma

-0.48

pg

-0.48

 proble

-0.47

ggle

-0.47

ggie

-0.47

POSITIVE LOGITS

cipled

0.58

 Soros

0.57

 consulate

0.56

isan

0.56

 Memor

0.53

 Benghazi

0.53

 ISIL

0.53

wagen

0.53

 Telegram

0.53

Lago

0.52

Activations Density 1.716%

phrases related to specific events, locations, and titles

key terms related to societal issues and cultural discussions

No Comments

No Known Activations